LiveWan 3.0 is live: 30-second clips, 1080p, native sound and dialogue. Try it now →

2.7
The multimodal Wan model

Wan 2.7
Alibaba's AI video model.

Wan 2.7 reads text, images, video clips and audio and returns a finished video: shots planned, camera moving, sound and dialogue already in. Generate text-to-video, image-to-video and reference-driven video in the browser.

Try Wan 2.7Available in the studio alongside the new Wan 3.0.
Overview

What is Wan 2.7?

Wan 2.7 is an AI video generation model from Alibaba's Tongyi Lab, the team behind the Wan (Tongyi Wanxiang) family. It is a multimodal model: you can prompt it with text, images, video clips and audio, and it returns a coherent video with camera direction, consistent characters and native sound.

Released in June 2026, Wan 2.7 is the most controllable model in the family: reference inputs, first-and-last-frame control and instruction-based video editing. It's available on Wan Studio alongside its successor, Wan 3.0. You get a simple prompt box, no downloads, and a free tier to start. Developers can also call it through the Wan API.

Features

Wan 2.7 features.

Prompting

Wan 2.7 speaks the language of film.

Camera moves, lighting setups, lens choices, emotional tone, write a Wan prompt the way a director would and 2.7 executes it. No settings panels, just words.

PromptSlow dolly-in on a street racer leaning on her car at night, neon rim light, rain just ended. She smirks and says "you're already too late." Cut to wide as the engine roars, bass-heavy score kicks in.
Getting started

How to use Wan 2.7.

  1. 01

    Sign in and open the studio

    Create an account and pick Wan 2.7 from the model selector, no waitlist, no invite code and nothing to download.

  2. 02

    Write a prompt and add references

    Describe the shot, the camera move and any dialogue in quotes. Optionally attach images and clips and tag them in the prompt, or switch to Edit mode with a base video.

  3. 03

    Generate and download

    Wan 2.7 returns a finished clip of up to 15 seconds at 720p or 1080p, with sound. Pick an aspect ratio for TikTok, Reels, Shorts or ads and export.

Release date

Wan 2.7 release date.

Alibaba released Wan 2.7 in June 2026. It followed Wan 2.5 (September 2025, the first Wan with native audio) and Wan 2.6, each step improving motion, prompt-following and audio quality.

Its successor, Wan 3.0, launched in September 2026 with 30-second clips and adaptive framing. On Wan Studio it's now the default model for every plan, and 2.7 remains available as the go-to for references and video editing.

Spec sheet

Wan 2.7 specifications.

Clip length
2–15 seconds
Resolution
720p · 1080p
Aspect ratios
16:9 · 9:16 · 4:3 · 3:4 · 1:1
Reference inputs
images · video clips · audio
Audio
native music, SFX, dialogue
Modes
T2V · I2V · references · video edit

Wan 3.0 is here.

Wan 3.0 doubles the runtime to 30 seconds, adds adaptive framing and sharper motion, on every plan, at no extra cost. It's live now.

Explore Wan 3.0
FAQ

Wan 2.7 FAQ.

Start creating with Wan 2.7.

References, video edit and native dialogue, see what Wan 2.7 does with your idea.

Make your first video